New Fauna from Loperot Contributesto the Understanding of Early MioceneCatarrhine Communities

Ari Grossman

Email agross@midwestern.edu

Cynthia Liutkus-Pierce

Benson Kyongo

Francis M’Kirera

Department of Anatomy, MidwesternUniversity, Glendale, Arizona, 85308 USA

School of Human Evolution and Social Change, Arizona StateUniversity, Tempe, Arizona, 85287 USA

Department of Geology, Appalachian State University, Boone, NorthCarolina, 28608 USA

Casting Department, National Museums of Kenya, Nairobi, Kenya

Department of Anatomy, Ross University School of Medicine, NorthBrunswick, New Jersey, 08902 USA

Abstract

The site of Loperot in West Turkana, Kenya, is usually assigned to theEarly Miocene. Recent discoveries at Loperot, including catarrhineprimates, led to a revision of its mammalian fauna. Our revision of thefauna at Loperot shows an unusual taxonomic composition of thecatarrhine community as well as several other unique mammalian taxa.Loperot shares two non-cercopithecoid catarrhine taxa with Early Miocene

1,2,*

3

4

5

1

2

3

4

5

sites near Lake Victoria, e.g., Songhor and the Hiwegi Formation ofRusinga Island, but Loperot shares a cercopithecoid, Noropithecus, withBuluk (Surgei Plateau, near Lake Chew Bahir). We use Simpson’s FaunalResemblance Index (Simpson’s FRI), a cluster analysis, and two partialMantel tests, to compare Loperot to 10 other localities in East Africarepresenting several time divisions within the Early and Middle Miocene.Simpson’s FRI of mammalian communities indicates that Loperot is mostsimilar in its taxonomic composition to the Hiwegi Formation of RusingaIsland, suggesting a similarity in age (≥18 Ma) that implies that Loperot isgeographically distant from its contemporaries, i.e., Hiwegi Formation ofRusinga Island, Koru, Songhor, and Napak, while at the same time olderthan other sites in West Turkana (Kalodirr and Moruorot). The clusteranalysis of the similarity indices of all the localities separates Loperot fromother Early Miocene sites in the study. Two partial Mantel tests show thatboth temporal distance and geographic distance between sites significantlyinfluence similarity of the mammalian community among sites. Thus,Loperot’s unique location in space and time may explain why it has anunusual catarrhine community and a number of unique taxa not seenelsewhere.

KeywordsCatarrhine primatesCluster analysisEarly MioceneHiwegi Formation of Rusinga IslandLoperotNon-cercopithecoidsPartial Mantel testsSimpson’s Faunal Resemblance Index

AQ1

Electronic supplementary material

The online version of this article (doi: 10.1007/s10764-014-9799-8 ) containssupplementary material, which is available to authorized users.

IntroductionCercopithecoidea is a diverse and successful clade that comprises themajority of living catarrhine primate species (Disotell 1996 ).Cercopithecoids have a wide geographic distribution over most of the OldWorld, encompassing a wide latitudinal gradient, and are found in diversehabitats such as tropical and subtropical forests, woodlands, savanna, andgrasslands (Jablonski and Frost 2010 ). By contrast, today, non-cercopithecoid catarrhines, represented by modern apes and humans, arerelatively taxon poor (Fleagle 19989). This pattern contrasts with the EarlyMiocene when non-cercopithecoid catarrhines were more taxonomicallydiverse than cercopithecoids (Fleagle 1998; Jablonski and Frost 2010 ) eventhough the two groups apparently began diverging by the Late Oligocene(Stevens et al. 2013 ). Moreover, although cercopithecoids are clearly wellestablished in Africa by the Early Miocene (Miller et al. 2009 ), they areusually rare elements in the mammalian community at that time (Jablonskiand Frost 2010 ).

AQ2

Catarrhine paleocommunities during the Early and Middle Miocene of Africa(Table I ) occupied a large diversity of habitats and exploited many niches,some quite different from those of modern apes and monkeys (Leakey et al.2011 ). Early and Middle Miocene monkeys are usually viewed as membersof the Victoriapithecidae, part of Cercopithecoidea, which also includesmodern Cercopithecidae (Jablonski and Frost 2010 ). Recent work (Miller etal. 2009 ) demonstrates increased diversity within the Victoriapithecidae (thesister family of modern Cercopithecidae). This diversity serves to highlightthe complexity of evolutionary history within the Cercopithecoidea. Early toMiddle Miocene non-cercopithecoid catarrhines are sometimes viewed asmembers of the Hominoidea (Michel 2014), or stem-Hominoidea (Stevens etal. 2013 ). Yet, some authorities prefer to place them in a broader radiationthat includes the Dendropithecoidea and Proconsuloidea (Harrison 2010 ) andexclude most from the Hominoidea. Nevertheless, a broad division existsbetween cercopithecoid monkeys and the non-cercopithecoid catarrhines. Yet,despite their abundance in Africa during the Early Miocene, we cannotpredict when apes and monkeys will be found together at a site, or when only

one group is more likely to be present.

Table I

African catarrhine paleocommunities used in this study: Age, geographic location, and habitat reconstructions

Site(coordinates) Age (Ma) Non-cercopithecoid

catarrhines

Early Miocene

Eastern Uganda

Moroto(2°31ʹ30.0ʹʹN, 34°46ʹ21.0ʹʹE)http://en.wikipedia.org/wiki/Mount_Moroto#cite_note-2

Aquitanian(20.6)(Gebo etal. 1997 ;cf. Pickfordand Mein2006 foranalternativeview)

Afropithecusturkanensis;Kogolepithecusmorotoensis;Micropithecus sp.

Napak(34°14ʹE; 02°05ʹN)(Bishop 1967 )

EarlyBurdigalian(18.5–20)(Pickfordet al.2010 )

Proconsul majorProconsul africanusDendropithecusmacinnesiLimnopithecuslegetetMicropithecusclarkiLomorupithecusharrisoni

Tinderet Sequence – Western Kenya (Lake Victoria)

Koru35°16ʹE; 00°09ʹS(Bishop 1967 )

EarlyBurdigalian(19–20)(Bishop etal. 1969 )

Proconsul majorProconsul africanusDendropithecusmacinnesiLimnopithecuslegetetKalepithecussonghorensisMicropithecusclarki

Songhor35°13ʹE; 00°02ʹS(Bishop 1967 )

19–20;EarlyBurdigalian(Bishop etal. 1969 )

Proconsul majorDendropithecusmacinnesiRangwapithecusgordoniLimnopithecusevansi

Kalepithecussonghorensis

Kisingiri Volcano – Western Kenya (Lake Victoria)

Rusinga – Hiwegi00°02ʹS; 35°13ʹE(Bishop 1967 )

EarlyBurdigalian(≥18 Ma)(Peppe etal. 2011 )

Proconsul heseloniProconsul nyanzaeDendropithecusmacinnesiLimnopithecuslegetetNyanzapithecusvancouveringorum

Rusinga – Kulu00°02ʹS; 35°13ʹE(Bishop 1967 )

LateBurdigalian(15–17)(Peppe etal. 2009 )

Proconsul heseloniProconsul nyanzaeDendropithecusmacinnesi

Turkana Region

Loperot(2°20ʹ0ʹʹN; 35°51ʹ0ʹʹE)

EarlyBurdigalian(ca. 19 Ma)(thisarticle)

LimnopithecuslegetetRangwapithecusgordoni

Kalodirr(3°20ʹN, 35°45ʹE)(Boschetto 1988 )

LateBurdigalian(16.8–17.5± 0.3 Ma)(Boschetto1988 )

AfropithecusturkanensisTurkanapithecuskalakolensisSimiolus enjiessi

Moruorot(3°17ʹN, 35°50ʹE)(Boschetto 1988 )

LateBurdigalian(16.8–17.5± 0.3 Ma)(Boschetto1988 )

AfropithecusturkanensisTurkanapithecuskalakolensisSimiolus enjiessi

Chew Bahir – Suregie plateau

Buluk(4°16ʹN, 36°36ʹE)(Harris and Watkins 1974 )

LateBurdigalian(>17.2)(McDougalandWatkins1985 )

AfropithecusturkanensisSimiolus enjiessi

North Africa

Wadi Moghara, Egypt(38°20ʹN, 28°30ʹE)(Approximated from Figure 1.1 in Hasan 2013 )

LateBurdigalian(17–18)(Miller1999 )

—

Jebel Zelten, Libya(28°00ʹN, 20°30ʹE)(Approximate location of Wadi Shatirat in map ofWessels 2003)

Early andMiddleMiocene inat leastthreedistincthorizons(Wessels etal. 2003 )

—

Middle Miocene

Western Kenya – Lake Victoria

Maboko Island(00°10ʹS; 34°36ʹ30ʹʹE)(Bishop 1967 )

Langhian(ca. 15 Ma)(Retallacket al.2002 )

EquatoriusafricanusMabokopithecusclarkiNyanzapithecuspickfordiMicropithecusleakeyorumLimnopithecusevansi

Fort Ternan(00°13ʹS; 35°21ʹE)(Bishop 1967 )

LateLanghianor earliestSerravalian(13.7 ±0.3–13.8 ±0.3)(Pickfordet al.2006 )

KenyapithecuswickeriSimiolus sp.Proconsul sp.

References for taxonomy: Cote 2008 ; Drake et al. 1988 ; Harrison 2010 ; Leakey et al. 2011al. 2009 ; Patel and Grossman 2006 ; Peppe et al. 2009 ; Pickford 2002 ; as well as personal observation.

References for habitat reconstructions: Andrews 1992 , 1996 ; Andrews and Van Couvering 1976 ; Andrews et al. 1979 , 1981 , 1997 ; Behrensmeyer et al. 2002 ; Cerling et al. 1991 , 19921970 ; Grossman 2008 ; Hill et al. 2013 ; Kappelman 1991 ; Kortlandt 1983 ; Leakey et al. Michel et al. 2014 ; Miller and Wood 2010 ; Nesbit Evans et al. 1981 ; Peppe et al. 2009 ; Pickford Pickford and Andrews 1981 ; Pickford and Mein 2006 ; Pickford et al. 2003 ; Retallack 19912002 ; Shipman 1986 ; Shipman et al. 1981 ; Ungar et al. 2012 ; Van Couvering and Van Couvering Gautier 1972 .

AQ3

It is well recognized that community composition is often allied withenvironmental conditions, so that communities in localities with differenthabitats, e.g. forests vs. woodlands vs. deserts, will differ in taxonomiccomposition, species richness, and abundance (Grossman 2008 ; Kamilar andBeaudrot 2013 ; Reed 1997 ). However, recognizing subtle environmentaldifferences among fossil sites is often difficult or unreliable because offactors such as taphonomic biases, lack of abiotic context (geology,sedimentology, etc.) or incomplete samples. Nevertheless, we can testwhether other factors, namely temporal differences and geographic distance,significantly affect similarities and differences in the taxonomic compositionof the mammalian communities among our study sites. Simply stated, localhabitats change over time through numerous local events combined withlarge-scale climatic and geologic events, all of which lead to effectivelychange community composition via extinction, speciation, migration, orimmigration (Preston 1960). This means we can expect sites that are closestin age to have more similar community composition as long as they are inroughly similar habitats. Similarly, as geographical distance increases, thedispersal propensity of species lessens, possibly because of an increasedchance of encountering geographic barriers or unsuitable habitat, whichaffects the similarity of community composition (Beaudrot and Marshall2011 ; Beaudrot et al. 2014 ; Kamilar 2009 ; Soininen et al. 2007 ). Thus,identifying the mammalian communities of Early and Middle Miocenecatarrhine-bearing sites and comparing them across time and space willprovide important information about the forces that shape the taxonomiccomposition of early catarrhine communities.

AQ4

To address questions about the effects of geographic distance and temporaldistance on community composition among Early and Middle Miocene siteswe need a sample of sites from different locations over similar time spans.However, Early Miocene catarrhine-bearing sites are unevenly distributed intime and space. Catarrhine-bearing sites of appropriate age are limited to EastAfrica with a few also in North Africa (including the Arabian Peninsula).This is complicated further by the incompleteness of faunal records for someof the sites. However, one site that can contribute important information toour understanding is Loperot, in West Turkana, in Kenya (Grossman 2013 ).Our efforts have led to recovery of a number of mammalian taxa, somepreviously unknown at the site. Among the fossils found are remains of bothcercopithecoid and non-cercopithecoid catarrhines. Unlike at most EarlyMiocene sites, cercopithecoid remains are much more abundant at Loperotthan non-cercopithecoids. However, to include Loperot in a comparisonamong Early and Middle Miocene sites we must first determine itsmammalian community and establish its age relative to that of other sites.

Thus, we aim to address the following questions:

1) What is the mammalian community, including the primates, of Loperot,and how does it compare with Early and Middle Miocene sites?

2) What is the age of the mammalian and primate community at Loperot asestimated from faunal comparisons with other Early and MiddleMiocene sites?

Once the first two questions are answered we can use this information toaddress questions about factors affecting the degree of similarity among Earlyand Middle Miocene mammalian and primate communities in their taxonomiccomposition. More specifically we ask:

3) Do temporal distance and geographic distance affect the composition ofthe mammalian communities at catarrhine-bearing localities of the Earlyand Middle Miocene in Africa?

MethodsDescription of the Loperot SiteThe fossil-bearing site of Loperot (2°20ʹ0ʹʹ North, 35°51ʹ0ʹʹ East) is located90 km south of the Lothidok range at the headwaters of the Kalabata River, atributary of the Keno River, which drains south–north to the southwesternshores of Lake Turkana (Fig. 1 ). This site is found within a larger area (2°–2°30ʹN, 35°30ʹ–36°E) with exposed Miocene rocks (Boschetto 1988 ;Boschetto et al. 1992 ; Joubert 1966 ). Previous research at Loperot identifiedan Early or Middle Miocene fauna that included monkeys and perhaps apes aswell as additional mammals (see Table II ).

Fig. 1

Map of Africa showing the location of the sites used in this study. 1) Loperot;2) Kalodirr and Moruorot; 3) Buluk 4) Moroto; 5) Napak; 6) Rusinga Island; 7)Songhor; 8) Koru; 9) Maboko; 10) Fort Ternan; 11) Wadi Moghara; 12) GebelZelten Circles = Early Miocene; Squares = Middle Miocene.

Table II

Taxonomic list of the fauna found in the Early Miocene fossiliferous deposits atLoperot

Primates indet

Victoripithecidae Pliohyracidae

Noropithecus sp. nov. cf. Meroehyrax batae

Proconsulidae Perrisodactyla

Rangwapithecus gordoni Rhinocerotidae

Family incertae sedis Chilotheridium pattersoni

Limnopithecus legetet cf. Brachypotherium sp.a

Rodentia Artiodactyla

Thryonomyidae Anthracotheriidae

Paraphiomys stromeri Brachyodus aequatorialis

Diamantomyidae Afromeryx cf. zelteni

Diamantomys leuderzi Tragulidae

Carnivora Dorcatherium pigotti

Felidae Dorcatherium chappuisi

Indet (small) Giraffidae

Creodonta cf. Canthumeryx syrtensis

Hyaenodontidae Suidae

cf. Hyainailouros cf. Kenyasus rusingensis

Indet (Medium-Small species) Ziphidae

Proboscidea indet

Deinotheriidae

Prodeinotherium hobleyi

Gomphotheriidae

Indet

Platybelodon sp.

cf. Archaeobelodon

Hyracoidea

Taxa published previously but not seen by authors.

Based on personal observations and the following references: Andrews 1978 ;Black 1978 ; Geraads 2010 ; Gingerich 2010; Guérin 2000; Harrison 1982 , 2010 ;Hoojier 1971; Lavocat 1978 ; Maglio 1969 ; Mead 1975 ; Pickford 1991 ; Sanderset al. 2010 ; Savage and Williamson 1978 ; Simons and Delson 1978 .

AQ5

In June 2012, we conducted a research expedition to Loperot, identifying fourlocalities (LpM1–LpM4) that yielded numerous vertebrate and invertebratefossil remains.

a

a

a

Regional Geologic History of LoperotThe Lokichar Basin Loperot sits on the western side of modern-day LakeTurkana, a rift lake along the East African Rift System (EARS) situatedbetween the Kenyan and Ethiopian domes (Feibel 2011 ; Morley et al.1999a ) (Fig. 2 ). As a result of Tertiary (as well as more recent) faulting,various different basins formed, including the Lokichar Basin, in which thesite of Loperot sits (Feibel 2011 ). The Lokichar Basin is a north–southtrending half graben that is ca. 60 km long by 30 km wide and is bound by aneast-dipping border fault on the west (Morley et al. 1999b ). The LokicharBasin is separated from the neighboring Kerio Basin by a topographic high ofgneissic basement (the Lokone Horst) (Hendrie et al. 1994 ; Morley et al.1999b ). Sediment infill of the Lokichar Basin is on the order of 7 km andthickens to the west near the border fault (Feibel 2011 ; Morley et al. 1999b ).Analysis of the basin sediments indicates Eocene through Late Miocene strata(Boschetto et al. 1992 ; Joubert 1966 ; Morley et al. 1999a ).

Fig. 2

Shaded relief map of the southwestern side of Lake Turkana and the location ofthree of the LpM sites visited in 2012. The sites sit to the west of the LokoneHorst (oblong topographic feature ca. 10 km to the east of the LpM sites thatstrikes NE-SW).

StratigraphyAQ6

Basin-fill sediments below the Auwerwer Basalts were known previously asthe Turkana Grits (Joubert 1966 ). These sediments are now subdivided intothe Lokone and Auwerwer Formations (Morley et al. 1999b ). The LokoneFormation (Paleogene–Early Miocene) consists of fluvio-deltaic, arkosicsandstones intercalated with two lacustrine shale units (the older LoperotShale Member and the younger Lokone Shale Member) (Morley et al.1999b ). Fluvial and lacustrine environments are confirmed by the presenceof crocodile, tortoise, oyster, and fish fossils within these sediments (Feibel2011 ; Joubert 1966 ). The Auwerwer Sandstone Formation (Middle Miocene)overlies the Lokone Member, and contains considerable volcaniclasticsediments including reworked tuffs (Morley et al. 1999b ). The AuwerwerBasalt directly overlies the Awerwer Sandstone Formation. It is ca. 300 mthick and gives an age of 12.5–10.7 Ma (Morley et al. 1999b ).

Fossil-Bearing Sediments Near LoperotThree primary sites (LpM2, LpM3, and LpM4) were identified as potentialfossil localities during 2012 fieldwork (Fig. 2 ), with LpM4 being the mostfossiliferous. At LpM4 we were able to find the location where M. Leakeyand her team found monkey remains in the past.

Sedimentology and Depositional EnvironmentsA brief description of the sedimentology at each fossil locality is provided inTable III . The presence of aquatic species, such as crocodiles, fish, andvarious invertebrates, coupled with the cross-bedding, climbing ripples, andfining-up sequences of the fossil-bearing quartz-rich sand units at all threesites indicate fluvial deposition. Thin, fining-up sequences found at LpM4suggest shallow, quick (braided) stream flow (with coarse bedload), at least atthat site. Red to green silts/sands above and below the fossil-bearing layersrepresent paleosols, as noted by the ped structure, rhizoconcretions, abundantplant microfossils, and the presence of gypsum and gibbsite. Preliminaryidentification of pollen grains we found at site LpM4 indicates an abundanceof grass (unpublished data). No lacustrine units (dark gray–black shales)were identified at any of the sites.

Table III

Summary of sedimentological features seen in samples from LpM sites, collected in 2012

Sample Color Grainsize Composition Texture Stuctures Environmental

interpretation

LpM2Unit1.1

Beige toyellow

Finesand tosilt

Quartz,feldspar,calcite,mafics, mica

Poorlysorted,coarseskewed

Fluvial

LpM2Unit1.2

RedMediumtocoarsesand

Quartz,calcite,gibbsite,mafics, mica

Poorlysorted

Peds,rhizoconcentrations,carbonate nodules

Paleosol/Floodplain

LpM3Unit2.1

BrownMediumtocoarsesand

Quartz,minorfeldspar

Poorlysorted

Contains bonefragments andinvertebrates

Fluvial

LpM3Unit2.2

Red/Green(mottled)

Finesand tosilt

Quartz,feldspar,calcite,gibbsite

Poorlysorted,coarseskewed

Peds? Paleosol/Floodplain

LpM3Unit3.1

BeigeMediumtocoarsesand

Quartz,feldspar,mafics

Moderatelysorted Rhizoconcentrations Fluvial (flow to

SW)

LpM3Unit3.2

RedMediumtocoarsesand

Quartz,feldspar,calcite,mafics, mica

Poorlysorted, fineskewed

Paleosol/Floodplain

LpM4Unit4.1

Beige toyellow

Fine tomediumsand

Quartz,feldspar,mafics

Poorlysorted

Bones, crossbeds,5–10 cm fining upsequences, grasspollen

Fluvial (possiblybraided streams)

LpM4Unit4.2

Red Finesand

Quartz,feldspar,mica, calcite,gibbsite,gypsum

Poorlysorted

Platy peds, gypsumcrystals, plantmicrofossils

Paleosol/Floodplain

The sedimentology therefore indicates a fluvial depositional environment,suitable (large enough, perennial) to sustain crocodiles, fish, and oysters.Paleosols, however, indicate periods of stability on the land surface, long

enough for the sediments (likely floodplain silts and fine sands) to alter tosoils. It is difficult to know at this time whether the change from a fluvialenvironment (the quartz sands) to the more stable land surface (paleosols)was due to a climate change, e.g., climate dries and the river ceases to flow,or more simply a change in the river’s course, e.g., anastomosing.

AgeRadiometric age dating of the basalt unit at LpM4 is currently in progress.Even so, the fossil-bearing units at all three primary sites are quartz- andfeldspar-rich sands (arkosic composition). This suggests that they belong tothe Lokone Formation and are Paleogene–Early Miocene in age (Morley etal. 1999b ). Baker et al. ( 1971 ) provide an age range of 18.0–16.2 Ma forbasalts at Loperot. Hooijer ( 1971 ) provides a radiometric age date of 18.0 ±0.9 Ma on a lava situated stratigraphically higher, e.g., younger, thanrhinoceros fossils at a Loperot location <2 km from the LpM3 site. However,Mead ( 1975 ) provides an approximate age of 17.1 Ma for Loperot. Boschettoet al. ( 1992 ) provide age determinations of 13.9 ± 0.2 Ma and 15.0 ± 0.2 Mafor Loperot but argue that these dates are too young because of Argon loss intheir samples. These reports, coupled with radiometric age dates of tuffaceousstrata to the north, led Brown and McDougall ( 2011 ) to suggest thatmammalian fossils found at Loperot are Early Miocene, while refraining fromconstraining Loperot within a specific time range.

Following our fieldwork at Loperot, we compiled a revised faunal list(Table II ). This list includes our discoveries of taxa previously unknownfrom Loperot combined with reanalysis of previously published materials. Tocompare the Loperot material to other sites in East Africa (Fig. 1 ), wecompare the mammalian assemblage of Loperot with the faunal assemblageof 10 other fossil sites (see Electronic Supplementary Material Table S I ).

Age Determination for Loperot Using Simpson’s FaunalResemblance IndexWe use Simpson’s Faunal Resemblance Index (FRI) to compare among alllocalities that have a taxonomically sufficient faunal list (Table IV ).Simpson’s FRI is defined as: FRI (%) = (Nc/ N1) × 100, where Nc is thenumber of identified taxa shared by two faunas, and N1 is the number of

identified taxa in the smaller of the two faunas (Simpson 1960 ). Simpson’sFRI conservatively measures the similarity among two assemblages bysimultaneously minimizing the use of samples of uneven size, and is verycommon in paleontological research (Holroyd and Ciochon 1994 ; Miller1999 ; Nakaya 1994 ; Tsubamoto et al. 2004 ). Also, when the taxonomic listsof faunal assemblages at different sites have large differences in size,Simpson’s FRI minimizes the effects of this difference (Miller 1999 ;Tsubamoto et al. 2004 ). We perform this analysis at the generic levelbecause, typically, genus-level data are more taxonomically stable and robustthan species-level data (Alroy 1996 ; Cifelli 1981 ; Tsubamoto et al. 2004 ).

Table IV

Results of the Simpson’s Faunal Resemblance Index examining faunal resemblance atthe genus level among a set of Early and Middle Miocene sites

Testing the Significance of Geographic Distance andTemporal Distance in Explaining Taxonomic CompositionDifferences Among Sites

Cluster Analysis

We use a hierarchical Unpaired Group Mean Average (UPGMA) clusteranalysis (Rohlf 1970 ) to determine whether the sites in our analysis formdiscrete groups. We performed the analysis by converting our similaritymatrix of 11 sites to a dissimilarity matrix (1 – FRI) using the genus-level

FRI values. We performed this analysis using the SPSS statistics package(IBM release 2009 ).

Partial Mantel Tests

We performed two partial Mantel tests (Mantel 1967 ; Ossi and Kamilar2006 ; Smouse et al. 1986 ) using the Vegan package (Oksanen et al. 2013 )for R (R Core Team 2014 ) to examine the relative importance of temporaldistance and geographic distance between sites. Age estimates and locationdata are provided in Table I . The first partial Mantel test looked at thecorrelation between taxonomic distance (converting the FRI values todissimilarity matrix 1 – FRI) and temporal distance while controlling forgeographic distance. The second looked at the correlation between taxonomicdistance and geographic distance while controlling for temporal distance.

Taphonomic and Collection BiasesTaphonomic and collecting biases affect the species composition of fossilassemblages as well as the resulting perceived structure of the community.We include in our study only fossil assemblages that include mammals of allsize classes and where we cannot a priori identify particular biases thatwould require exclusion of these sites from the analysis. Therefore, weexclude the Early Miocene sites of Buluk, Wadi Moghara, and Gebel Zelten,from this analysis even though these important sites preserve fossil monkeys.Buluk (Anemone et al. 2005 ) and Wadi Moghara (Miller 1999 ) have norodents or other micromammals published. This biases the composition of theassemblages sufficiently to exclude these sites from the Simson’s FRIanalysis. Gebel Zelten preserves a large number of micromammals thatindicate the assemblage is a time-average of at least four different time zones(Wessels et al. 2003 ), making it unusable. We exclude an important MiddleMiocene catarrhine-bearing site, Kipsaramon, from our analysis. Although asmall number of recently published specimens from Kipsaramon assigned toVictoriapithecidae genus et species indet. are from sediments estimated to be15.83–15.59 Ma (Gilbert et al. 2010 ), most of the cercopithecoid materialsdescribed from Kipsaramon were assigned to cf. Noropithecus (Miller et al.2009 ) and are assigned an approximate age of ca. 14.5 Ma (Pickford andKunimatsu, 2005 ) further confounding the issue. At present it is unclearwhether these two different articles represent parts of a single primate

community or several different ones. In addition, fauna from Kipsaramon arenot fully described limiting utility of the site for our analysis.

ResultsWe present the revised list of taxa found at Loperot (Table II ) as a singlefauna even though the new fossils we discovered in 2012 are from fourdifferent loci within the larger Loperot site and other fossils may be fromslightly different localities. We feel confident that they form a singleassemblage because there is substantial overlap between the fauna found atmost localities. Overall, the fauna from Loperot are primarily taxa that areknown elsewhere in the Early Miocene of East Africa. Still, as mentionedpreviously, some taxa are known only from Loperot.

Overall, Loperot is most similar to the Hiwegi Formation of Rusinga Island(Table IV ; FRI = 63.2). Next, Loperot is most similar to Napak (FRI = 47.4).Following that, Loperot is most similar to Kalodirr and Moruorot, the othertwo sites in West Turkana.

The Early and Middle Miocene are separate clusters (Fig. 3 ). The EarlyMiocene cluster is split into two distinct clusters. One cluster includesLoperot, Kalodirr, and Moruorot. The second includes all the other EarlyMiocene sites in our study. In the first cluster, Loperot is quite distantlylinked to a distinct cluster that includes Kalodirr and Moruorot. In the secondEarly Miocene cluster, the two Rusinga assemblages form a distinct cluster,as do Koru and Songhor. Napak is nestled in a cluster with Koru andSonghor. Moroto is distantly linked to all the sites in the second cluster. TheMiddle Miocene sites, although forming a distinct cluster, are not closelylinked either.

Fig. 3

A dendrogram showing the results of an using Unpaired Group Mean Average(UPGMA) cluster analysis of the 1-FRI dissimilarity matrix for sites used inthe FRI analysis.LP=Loperot; WK=Kalodirr; MO=Moruorot; RU-H=Rusinga-Hiwegi Fm.; RU-K=Rusinga-Kulu Fm., SO=Songhor; KO=Koru; NP=Napak;MR=Moroto; MB=Maboko; FT=Fort Ternan.

AQ7

In the first partial Mantel test there is a significant correlation betweentaxonomic distance and temporal distance while controlling for geographicdistance (Mantel statistic r: 0.5468; P = 0.001). In the second partial Manteltest there is a significant correlation between taxonomic distance andgeographic distance while controlling for temporal distance (Mantel statisticr: 0.3105; P = 0.022).

DiscussionAge of Loperot as Indicated by the FaunaAs demonstrated by the Simpson’s FRI, Loperot’s best match (FRI = 63.2) isto the Hiwegi Formation of Rusinga Island (Rusinga–Hiwegi in Table IV ).The Rusinga–Hiwegi fauna is the largest and many other sites also have FRIvalues in the 60–70 range with Rusinga, e.g., Kalodirr, Moruorot, Songhor,Koru, and Napak. This could suggest that the Rusinga–Hiwegi faunalassemblage may skew the results of this analysis. However, Kalodirr andMoruorot are most similar to each other (FRI = 91.3), Songhor and Koru aremost similar to each other (FRI = 91.2), and Napak is also most similar toKoru (FRI = 73.5). In fact, only Loperot and Rusinga-Kulu are most similarto Rusinga–Hiwegi. However, Loperot and Rusinga–Kulu are not verysimilar (FRI = 36.8). Therefore, it is unlikely that the Hiwegi Formationfauna from Rusinga dominates the FRI analysis enough to hide real patternsin the data.

The results of the FRI analysis support the placement of Loperot in the Early

Miocene rather than the Middle Miocene, in accordance with Brown andMcDougall ( 2011 ). Until recently, the Hiwegi Formation of Rusinga wasdated to 17.8 Ma (Drake et al. 1988 ). Peppe et al. ( 2011 ) provide newinformation that suggests that the Hiwegi Formation is in fact older (≥18.0Ma based on text and Fig. 2 of Michel et al. 2014 ) than previously thought,but still not as old as Songhor and Koru (ca. 19.6 Ma; Bishop et al. 1969 ;Hill et al. 2013 ) and Napak (ca. 19.6; Bishop et al. 1969 ; Senut et al. 2000 ).The latter sites may be age equivalent with the older Wayando Formation ofRusinga Island and Mfangano (Peppe et al. 2011 ).

It is therefore intriguing that Loperot shares two taxa with Songhor to theexclusion of Rusinga–Hiwegi or any other sites. The primates at Loperotinclude Rangwapithecus gordoni, previously only known from Songhor (Hillet al. 2013 ) and the nearby and age-equivalent Lower Kapurtay locality(Cote et al. 2014 ). Both sites are within the Kapurtay AgglomeratesFormation of the Tinderet Sequence (Pickford and Andrews 1981 ), and allthe fauna found at Lower Kapurtay are also known from Songhor (Cote et al.2014 ). Finding Rangwapithecus gordoni at Loperot is a remarkable rangeextension for this species. The other taxon that Loperot shares only withSonghor is a proboscidean d4 (KNM-LP 53749: cf. Archaeobelodon inTable II ) that is almost identical to KNM-SO 1237 from Songhor. Both thesespecimen are more primitive than any amebelodontine proboscideanscurrently known from Rusinga–Hiwegi. This is evident by such features asthinner enamel, lack of posterior accessory cuspules, and relatively smallsize. Both of these taxa suggest that Loperot may be older than Rusinga–Hiwegi, although more data are needed to test this idea. At Loperot, there areno taxa currently recognized that indicate an age younger than 18 Ma, whichis in agreement with the fauna and the radiometric ages published by Hooijer( 1971 ). Therefore, we estimate the age of Loperot as older than 18 Ma andperhaps closer to 19 Ma.

Effects of Temporal Distance and Geographic Distance onCommunity CompositionAlthough a cursory look at the dendrogram presented in Fig. 3 may suggest aregional grouping pattern, this may be true only for sites that are very closegeographically, i.e., the Rusinga sites from two different formations, Koru

and Songhor, and Kalodirr and Moruorot. Even though Napak isgeographically closest to Moroto, Napak is nestled in a cluster with Songhorand Koru, and then within a larger cluster that includes the Rusinga localities.Both Moroto and Loperot are linked only distantly to the other sites in thestudy. In fact, looking at the sites in Turkana and the sites from Uganda, itappears that temporal distance is more likely to explain the distances inlinkage between sites from different times. Our estimates for Loperot indicatethat it is older than Kalodirr and Moruorot, while Moroto is older than Napak.Napak actually clusters closer to contemporaneous localities (Songhor andKoru) that are geographically more distant than Moroto.

Given the results of the cluster analysis, combined with the unusual primatecommunity and unique fauna at Loperot, we wanted to determine whatfactors are of primary importance in determining the composition mammaliancommunity of Early Miocene sites. Habitat reconstructions for Loperot arequite preliminary, so we examined two other factors that may contribute todetermining the similarity in the composition of mammalian communities:temporal distance and geographical distance. The partial Mantel tests indicatethat both time and geographic location have an effect on taxonomic distance.However, at least in our data, temporal differences are more important.Therefore, we divide the Early Miocene sites in our study into threesubperiods to allow for more detailed examination. Moroto is the only sitewithin the Aquitanian (23.03–20.44 Ma; Cohen et al. 2013 ). We can dividethe Burdigalian (20.44–15.97 Ma; Cohen et al. 2013 ) into Early Burdigalian(≥ca. 18 Ma; Rusinga–Hiwegi, Napak, Koru, Songhor, and Loperot) and LateBurdigalian (ca. 18 Ma–15.97 Ma; Kalodirr, Moruorot, Rusinga–Kulu).Maboko and Fort Ternan are both Middle Miocene sites but interestinglyrepresent the two ends of the Langhian (15.97–13.82 Ma; Cohen et al. 2013 ).

Moroto is the oldest site and is indeed quite separate in its faunal compositionfrom the other Early Miocene localities (highest FRI = 51.7 with Napak andSonghor). Koru, Songhor, and Napak are contemporaneous (ca. 19.5 Ma,Bishop 1969) and indeed we can see how they form a distinct group in thecluster analysis (Fig. 3 ). Within this cluster we see that Napak,geographically more distant, is also separate from a cluster that includes onlyKoru and Songhor.

Given the greater correlation and significance of temporal distance weexpected Rusinga–Hiwegi and Loperot to cluster together and nestle within alarger cluster of early Burdigalian sites, while we expected Rusinga–Kulu tocluster with Kalodirr and Moruorot in a Late Burdigalian cluster. Indeed,Moruorot and Kalodirr cluster together, but the two Rusinga assemblages andLoperot do not cluster as expected. The faunas of the Hiwegi Formation andthe Kulu Formation are very similar, as evidenced by their low linkagedistance (Fig. 3 ), while Loperot is almost as distant from Kalodirr andMoruorot as Moroto is distant from the two Rusinga assemblages (seeFig. 3 ).

Comparing the two Rusinga assemblages, only one genus, Turkanatheriumacutirostratum, is reportedly present at Kulu (Peppe et al. 2009 ) but is notpresent in the Hiwegi assemblage. However, Geraads ( 2010 ) argues thatTurkanatherium acutirostratum cannot be identified anywhere but Moruorot(he had no access to the Kalodirr fossils at that time), which would then makethe two Rusinga assemblages identical at the genus level. The similaritiesamong the two sites may be a result of sampling size (Kulu = 26 genera;Hiwegi = 68 genera), but may also be the result of historically treating themany localities at Rusinga Island as a single time-averaged fauna (Michel etal. 2014). It seems unlikely that two faunal assemblages a million or moreyears apart (Peppe et al. 2009 , 2011 ) will be identical and perhaps there is aneed for reanalysis of the Rusinga faunas, especially given the importance ofthese localities to understanding of the Early Miocene, as demonstrated byMichel et al. ( 2014 ).

The small size of the Loperot assemblage (19 genera) may affect the analysis.However, despite being most similar to Rusinga–Hiwegi in its FRI analysis(SI = 63.2), Loperot does not cluster closely to Rusinga–Hiwegi or any othersites. It is only distantly clustered with other sites in Turkana. AlthoughLoperot is most similar to Rusinga–Hiwegi, other sites such as Rusinga–Kulu(SI = 96.2), Songhor (SI = 66.7), and others are even more similar toRusinga–Hiwegi. This affects the cluster analysis and provides a likelyexplanation to why Loperot is not close to Rusinga–Hiwegi in the clusteranalysis. Loperot clustering with Kalodirr and Moruorot is likely also theresult of other sites being more similar to each other and not any indication of

much similarity among the West Turkana sites. This is reflected in the verylarge linkage distance of Loeprot from Kalodirr/Moruorot. Loperot differsfrom other early Burdigalian sites in its geographic location and from othersites in West Turkana by its age. As both temporal differences andgeographic differences affect faunal composition, Loperot’s linkage distanceon the cluster analysis from the rest of the Early Miocene sites may very wellbe a real phenomenon resulting from its unique interaction of geography andtime.

Composition of Catarrhine Communities in Space andTimeLoperot shares its non-cercopithecoid catarrhines with Early Burdigalian sites(Songhor, Rusinga–Hiwegi, Koru, and Napak) rather than Late Burdigaliansites (Kalodirr, Moruorot, and Hiwegi–Kulu), even though Kalodirr andMoruorot are geographically closer to Loperot than any other sites. This is inaccordance with the greater influence of temporal difference on mammaliancommunity structure indicated by the partial Mantel tests. More importantly,this strongly suggests that the Afropithecus–Turkanapihtecus–Simioluscatarrhine community present in other younger Early Miocene sites in LakeTurkana (Leakey et al. 2011 ) very likely replaced a catarrhine communitythat primarily comprised taxa shared with other older sites also found in otherregions.

In addition to the cercopithecoid specimens at Loperot, a single upper molarfrom Napak (either M1 or M2), UMP 62-21, was described (although notillustrated) with damaged mesial and buccal margins but intact cusps(Pilbeam and Walker 1968). A frontal was published together with the toothbut has since been assigned to Micropithecus clarki (Fleagle and Simons1978 ). It is possible that reanalysis of the molar may remove it from theCercopithecoidea as well. At Moroto, a right lower canine and the lower p3and p4 of a single individual were assigned to Prohylobates macinnesi(Pickford et al. 2003 ). A recent revision by Miller et al ( 2009 ) erected anumber of new genera and species and placed this specimen inVictoriapithecus macinnesi. The cercopithecoid remains from Moroto wouldalso benefit from reexamination, particularly given the large temporal gapbetween Moroto and Maboko. Although Loperot may share the presence of

cercopithecoids with Moroto, at present these differ generically and the twosites do not share any non-cercopithecoid catarrhine taxa.

AQ8

Cercopithecoids in our studied sites do not follow the results of the partialMantel tests. Loperot does not share Noropithecus sp. with any of the sites inthe study. However, monkeys are known from three Late Burdigalian sitesthat we excluded from the study owing to taphonomic biases: Buluk (16.4 ±0.2–17.4 ± 1.6 Ma: McDougal and Watkins 2006), Wadi Moghara (17–18Ma: Miller 1999 ), and Gebel Zelten 18–15 Ma (Wessels et al. 2003 ). Theonly catarrhines currently known from the North African sites, WadiMoghara and Gebel Zelten, are cercopithecoids, Prohylobates tandyi andZaltanpihtecus simonsi respectively (Miller et al. 2009 ), different fromNoropithecus sp. from Loperot. Interestingly, Noropithecus was erected forthe cercopithecoid found at Buluk (Miller et al. 2009 ). Noropithecus differsfrom all other cercopithecoids in having more bunodont lower molar cusps,and greater degree of molar flare due to mesial and distal cusp tips beingmore closely approximated (Miller et al. 2009 ). A formal description of theLoperot cercopithecoid material is outside the scope of this article and is thesubject of an ongoing study; still, we place the monkey from Loperot in thegenus Noropithecus, albeit a different species than N. bulukensis from Buluk(see Table I ) because it is similar to N. bulukensis in its bunodont lowermolar cusps and close approximation of mesial and distal cusps, leading to ahigh degree of molar flare. However, Buluk shares non-cercopithecoidcatarrhines with Kalodirr and Moruorot and not Early Burdigalian sites. Twoother sites may also have Noropithecus present. Nabwal (<17.2 Ma; Fleagleet al. 1997 ) preserves cf. Noropithecus fleaglei, and Kipsaraman (14.5 Ma;Pickford, 1981 ) preserves cf. N. kipsaramanensis (Miller et al. 2009 ;Pickford and Kunimatsu 2005 ). Whether these are two species ofNoropithecus or not, they are apparently part of a “Noropihtecus” group.Loperot may well represent the oldest member of this group.

AQ9

During the Middle Miocene Victoriapithecus macinnesi is very well known atMaboko, (Benefit 1999 ), but is not present at Fort Ternan. Miller et al.( 2009 ) recognize Victoriapithecus macinnesi at three Early Miocene sites:Napak, Moroto, in Uganda, and Loperot in Kenya, and several Middle

Miocene sites: Maboko Island, Majiwa, Nachola, Nyakach, Ngorora, andOmbo, all in Kenya. As mentioned previously, we reassign the material fromLoperot to Noropithecus sp., and argue that the very minimal material fromMoroto (one canine and two premolars from a single individual), and singleincomplete upper molar from Napak, require reanalysis. Therefore, wesuggest that it is currently best to treat Victoriapithecus macinnesi as aMiddle Miocene taxon only. It is important to note that Noropithecus mayhave survived into the Middle Miocene as cf. N. kipsaramenssis.

At present it is difficult to evaluate why some sites preserve monkeys alongwith apes, whereas most preserve only one or the other. In the case of the twoNorth African sites, one can postulate some barrier to the migration of non-cercopithecoids, as only cercopithecoids are identified in North Africa or theArabian Peninsula before the early Middle Miocene when Heliopithecus isfound at Ad-Dabtia in Saudi Arabia (Andrews and Martin 1987 ; Andrews etal. 1978 ; Harrison 2010 ). However, no such barrier has been demonstrated.

Temporal differences account for much of the difference in communitycomposition between sites. Loperot shares non-cercopithecoids with thesimilar-age Hiwegi fauna of Rusinga (Michel et al. 2004) and with maybeeven older Songhor and Lower Kapurtay (Cote et al. 2014 ). The presence ofa monkey at Loperot may be the result of geographical differences in thedistribution of monkeys, perhaps also suggested by the lack of any monkeysin any sites near Lake Victoria. However, such an idea would require a muchgreater sampling of catarrhine habitats than is currently available. Thedifferences between the catarrhine communities of the Early Burdigalian andLate Burdigalian sites in East Africa are in agreement with the combinedeffect of both temporal difference and geographical difference.

ConclusionsOur analyses indicate that both time differences and spatial distance affect thesimilarity of community composition in Early Miocene sites. Our studyshows that catarrhine communities generally follow that pattern. Our resultsindicate possible turnover in catarrhine communities over time. Our resultsalso indicate that geographic differences also played a role in differentiatingmammalian and catarrhine communities. Further studies are necessary to

determine if Loperot is unusual in some aspects of its ecology orenvironments to explain why certain taxa and not others are shared amongLoperot and other sites, and why it shares taxa with sites that incorporate alarge range of time.

AcknowledgmentsFunding for this project was provided by The Leakey Foundation.Midwestern University and Appalachian State University provided logisticalsupport and additional funds. We thank the guest editors, K. Reed, J.Kamilar, and L. Beaudrot, for the invitation to participate in the symposiumthat led to this issue. We are grateful to the Kenyan government and NationalMuseums of Kenya for facilitating our research. We are especially grateful toDrs. E. Mbua and F. Manthi for project support. Thanks also to S. Longoria,T. Moru, and J. Ekeno. The people of Loperot deserve our special thanks fortheir friendship and assistance while in their land. This manuscript benefittedgreatly from reviews by K. McNulty and two anonymous reviewers.

Electronic supplementary materialBelow is the link to the electronic supplementary material.

ESM 1

(DOCX 83 kb)

References

Alroy, J. (1996). Constant extinction, constrained diversification, anduncoordinated stasis in North American mammals. Palaeogeography,Palaeoclimatology, Palaeoecology, 127(1), 285–311.

Andrews, P. (1978). A revision of the Miocene Hominoidea of EastAfrica. Bulletin of the British Museum of Natural History (Geology), 30,85–224.

Andrews, P. (1992). Community evolution in forest habitats. Journal of

Human Evolution, 22, 423–438.

Andrews, P. (1996). Palaeoecology and hominoid Paleoenvironments.Biological Reviews of the Cambridge Philosophical Society, 71, 257–300.

Andrews, P., Begun, D. R., & Zylstra, M. (1997). Interrelationshipsbetween functional morphology and paleoenvironments in Miocenehominoids. In D. R. Begun, C. V. Ward, & M. D. Rose (Eds.), Function,phylogeny, and fossils: Miocene hominoid evolution and adaptations (pp.29–57). New York: Plenum Press.

Andrews, P., Hamilton, W. R., & Whybrow, P. J. (1978). Dryopithecinesfrom the Miocene of Saudi Arabia. Nature, 274, 249–251.

Andrews, P., Lord, J. M., & Nesbit Evans, E. M. (1979). Patterns ofecological diversity in fossil and modern mammalian faunas. BiologicaJournal of the Linnean Society, 11, 177–205.

Andrews, P. J., & Martin, L. (1987). The phyletic position of the AdDabtiyah hominoid. Bulletin of the British Museum of Natural History(Geology), 41, 383–393.

Andrews, P., Meyer, G. E., Pilbeam, D. R., Van Couvering, J. A., & VanCouvering, J. A. H. (1981). The Miocene fossil beds of Maboko Island,Kenya: Geology, age, taphonomy and palaentology. Journal of HumanEvolution, 10, 35–48.

Andrews, P., & Van Couvering, J. A. H. (1975). Palaeoenvironments inthe East African Miocene. In F. S. Szalay (Ed.), Approaches to primatepaleobiology (pp. 62–103). Basel: Karger.

Andrews, P., & Walker, A. (1976). The primate and other fauna from FortTernan, Kenya. In G. L. Isaac & E. R. McCown (Eds.), Human origins:Louis Leakey and the East African evidence (pp. 279–306). Menlo Park,CA: W. A. Benjamin.

Anemone, R. L., Grossman, A., Miller, E. R., & Watkins, R. T. (2005).

Biochronology and paleoecology of the Buluk fauna, Early Miocene ofnorthern Kenya. Journal of Vertebrate Paleontology, 25(S3), 32A.

Baker, B. H., Williams, L. A. J., Miller, J. A., & Fitch, F. J. (1971).Sequence and geochronology of the Kenya rift volcanics. Tectonophysics,11, 191–215.

Beaudrot, L., Kamilar, J. M., Marshall, A. J., & Reed, K. E. (2014).African primate assemblages exhibit a latitudinal gradient in dispersallimitation. International Journal of Primatology DOI.10.1007/s10764-014-9773-5

AQ10

Beaudrot, L., & Marshall, A. J. (2011). Primate communities arestructured more by dispersal limitation than by niches. Journal of AnimalEcology, 80, 332–341.

Behrensmeyer, A. K., Deino, A. L., Hill, A., Kingston, J. D., & Saunders,J. J. (2002). Geology and geochronology of the Middle MioceneKipsaramon site complex, Muruyur Beds, Tugen Hills, Kenya. Journal ofHuman Evolution, 42, 11–38.

Benefit, B. R. (1999). Victoriapithecus: The key to Old World monkeyand catarrhine origins. Evolutionary Anthropology, 7, 155–174.

Bishop, W. W. (1967). The later Tertiary in East Africa-volcanics,sediments, and faunal inventory. In W. W. Bishop & J. D. Clark (Eds.),Background to evolution in Africa (pp. 31–56). Chicago: University ofChicago Press.

Bishop, W. W., Miller, J. A., & Fitch, F. W. (1969). New potassium-argonage determinations relevant to the Miocene fossil mammal sequence inEast Africa. American Journal of Science, 267, 669–699.

Black, C. C. (1978). Anthracotheridae. In V. M. Maglio & H. B. S. Cooke(Eds.), Evolution of African mammals (pp. 423–434). Cambridge, MA:Harvard University Press.

Boschetto, H. B. (1988). Geology of the Lothidok Range. Master’s thesis,University of Utah.

Boschetto, H. B., Brown, F. H., & McDougall, I. (1992). Stratigraphy ofthe Lothidok Range, northern Kenya, and K/Ar ages of its Mioceneprimates. Journal of Human Evolution, 22, 47–71.

Brown, F. H., & McDougall, I. (2011). Geochronology of the Turkanadepression of northern Kenya and southern Ethiopia. EvolutionaryAnthropology, 20, 217–227.

Cerling, T. E., Harris, J. M., Ambrose, S. H., Leakey, M. G., & Solounias,N. (1997). Dietary and environmental reconstruction with stable isotopeanalyses of herbivore tooth enamel from the Miocene locality of FortTernan, Kenya. Journal of Human Evolution, 33, 635–650.

Cerling, T. E., Kappelman, J., Quade, J., Ambrose, S. H., Sikes, N. E., &Andrews, P. (1992). Reply to Comment on the paleoenvironment ofKenyapithecus at Fort Ternan. Journal of Human Evolution, 23, 371–377.

Cerling, T. E., Quade, J., Ambrose, S. H., & Sikes, N. E. (1991). Fossilsoils, grasses, and carbon isotopes from Fort Ternan, Kenya: Grassland orwoodland? Journal of Human Evolution, 21, 295–306.

Cifelli, R. (1981). Patterns of evolution among the Artiodacyla andPerissodactyla (Mammalia). Evolution, 35, 433–440.

Cohen. K. M., Finney, S. .C, Gibbard, P. L., & Fan J-X. (2013; updated).The ICS International Chronostratigraphic Chart. Episodes, 36:199-204.Cohen, K. M., Finney, S., & Gibbard, P. L. (2013). InternationalChronostratigraphic Chart. International Commission on Stratigraphy

AQ11

Cote, S. M. (2008). Sampling and ecology in three Early Miocenecatarrhine assemblages from East Africa. Ph.D. dissertation, HarvardUniversity.

Cote, S., Malit, N., & Nengo, I. (2014). Additional mandibles ofRangwapithecus gordoni, an Early Miocene catarrhine from the Tinderetlocalities of Western Kenya. American Journal of Physical Anthropology,153, 341–352.

Disotell, T. R. (1996). The phylogeny of Old World monkeys.Evolutionary Anthropology, 5, 18–24.

Drake, R. E., Vancouvering, J. A., Pickford, M. H., Curtis, G. H., &Harris, J. A. (1988). New chronology for the Early Miocene faunas ofKisingiri, western Kenya. Journal of the Geological Society, London, 145,479–491.

Feibel, C. S. (2011). A geologic history of the Turkana Basin.Evolutionary Anthropology, 20, 206–216.

Fleagle JG. 1999. Primate Adaptation and Evolution. San Diego:Academic Press.

Fl

Fleagle, J. G., Bown, T. M., Harris, J. M., Watkins, R. W., & Leakey, M.G. (1997). Fossil monkeys from northern Kenya. American Journal ofPhysical Anthropology, S24, 111.

Fleagle, J. G., & Simons, E. L. (1978). Micropithecus clarki, a small apefrom the Miocene of Uganda. American Journal of Physical Anthropology,49, 427–440.

Gebo, D. L., MacLatchy, L., Kityo, R., Deino, A., Kingston, J., &Pilbeam, D. (1997). A hominoid genus from the Early Miocene of Uganda.Science, 276, 401–404.

Gentry, A. W. (1970). The Bovidae (Mammalia) of the Fort Ternan fossil

fauna. Fossil Vertebrates of Africa, 2, 243–323.

Geraads, D. (2010). Rhinocerotidae. In L. Werdelin & W. J. Sanders(Eds.), Cenozoic mammals of Africa (pp. 675–689). Berkeley: Universityof California Press.

Gilbert, C. C., Goble, E. D., & Hill, A. (2010). Miocene Cercopithecoideafrom the Tugen Hills, Kenya. Journal of Human Evolution, 59, 465–483.

Gingerich, P. D. (2010). Cetacea. In L. Werdelin & W. J. Sanders (Eds.),Cenozoic mammals of Africa (pp. 873–899). Berkeley: University ofCalifornia Press.

AQ12

Grossman, A. (2008). Ecological and morphological diversity incatarrhine primates from the Miocene of Africa. Ph.D. dissertation, StonyBrook: Stony Brook University.

Grossman, A. (2013). Primate paleocommunities in the Early Miocene ofAfrica: Why are apes and monkeys so rarely found together? AmericanJournal of Physical Anthropology, 150, 138.

Harris, J., & Watkins, R. T. (1974). New Early Miocene vertebrate localitynear Lake Rudolf, Kenya. Nature, 252, 576–577.

Harrison, T. (1982). Small bodied apes from the Miocene of East Africa.Ph.D. dissertation, University College London.

Harrison, T. (2010). Dendropithecoidea, Proconsuloidea, and Hominoidea.In L. Werdelin & W. J. Sanders (Eds.), Cenozoic mammals of Africa (pp.429–469). Berkeley: University of California Press.

Hasan, S. M. (2013). Sequence stratigraphy of the Lower Miocene MoghraFormation in the Qattara Depression, North Western Desert, Egypt. NewYork: Springer Science+Business Media.

Hendrie, D. B., Kusznir, N. J., Morley, C. K., & Ebinger, C. J. (1994).

Cenozoic extension in northern Kenya: A quantitative model of rift basindevelopment in the Turkana area. Tectonophysics, 236, 409–438.

Hill, A., Nengo, I. O., & Rossie, J. B. (2013). A Rangwapithecus gordonimandible from the Early Miocene site of Songhor, Kenya. Journal ofHuman Evolution, 65, 490–500.

Holroyd, P. A., & Ciochon, R. L. (1994). Relative ages of Eocene primate-bearing deposits of Asia. In J. G. Fleagle & R. F. Kay (Eds.), Anthropoidorigins (pp. 123–141). New York: Plenum Press.

Hooijer, D. A. (1971). A new rhinoceros from the late Miocene ofLoperot, Turkana District, Kenya. Bulletin of the Museum of ComparativeZoology, 142, 339–392.

Jablonski, N., & Frost, S. (2010). Cercopithecoidea. In L. Werdelin & W.J. Sanders (Eds.), Cenozoic mammals of Africa (pp. 393–428). Berkeley:University of California Press.

Joubert, P. (1966). Geology of the Loperot area. Degree Sheet 18, SEQuarter. No. 74. Ministry of Natural Resources, Geological Survey ofKenya.

Kamilar, J. M. (2009). Environmental and geographic correlates of thetaxonomic structure of primate communities. American Journal ofPhysical Anthropology, 139, 382–393.

Kamilar, J. M., & Beaudrot, L. (2013). Understanding primatecommunities: Recent developments and future directions. EvolutionaryAnthropology, 22, 174–185.

Kappelman, J. (1991). The paleoenvironment of Kenyapithecus at FortTernan. Journal of Human Evolution, 20, 95–129.

Kortlandt, A. (1983). Facts and fallacies concerning Miocene ape habitats.In R. Ciochon & R. Corruccini (Eds.), New interpretations of ape andhuman ancestry (pp. 465–514). New York: Plenum Press.

Lavocat, R. (1978). Rodentia and Lagomorpha. In V. M. Maglio & H. B.S. Cooke (Eds.), Evolution of African mammals (pp. 69–89). Cambridge,MA: Harvard University Press.

Leakey, M., Grossman, A., Gutiérrez, M., & Fleagle, J. G. (2011). Faunalchange in the Turkana Basin during the Late Oligocene and Miocene.Evolutionary Anthropology, 20, 238–253.

Madden, C. T. (1972). Miocene mammals, stratigraphy and environmentof Muruarot Hill, Kenya. PaleoBios, No. 14.

Maglio, V. J. (1969). A shovel-tusked gomphothere from the Miocene ofKenya. Breviora, 310, 1–10.

Mantel, N. (1967). The detection of disease clustering and a generalizedregression approach. Cancer Research, 27, 209–220.

McDougall, I., & Watkins, R. T. (1985). Age of hominoid-bearingsequence at Buluk, northern Kenya. Nature, 318, 175–178.

McDougall, I., & Watkins, R. T. (2006) Geochronology of the NabwalHills: a record of earliest magmatism in the northern Kenyan Rift Valley.Geological Magazine, 143:25-39.

Mead, J. G. (1975). A fossil beaked whale (Cetacea: Ziphiidae) from theMiocene of Kenya. Journal of Paleontology, 49, 745–751.

Michel, L. A, Peppe, D. J., Lutz, J. A., Driese, S. G., Dunsworth, H. M.,Harcourt-Smith, W. E., Horner, W. H., Lehmann, T., Nightingale, S., &McNulty, K. P. (2014). Remnants of an ancient forest provide ecologicalcontext for Early Miocene fossil apes. Nature Communications, 5, doi:10.1038/ncomms4236 .

Miller, E., Benefit, B., McCrossin, M., Plavcan, J., Leakey, M., El-Barkooky, A., Hamdan, M., Abdel Gawad, M., Hassan, S., & Simons, E.(2009). Systematics of early and middle Miocene Old World monkeys.

Journal of Human Evolution, 57, 195–211.

Miller, E. R. (1999). Faunal correlation of Wadi Moghara, Egypt:Implications for the age of Prohylobates tandyi. Journal of HumanEvolution, 36, 519–533.

Miller, E. R., & Wood, A. (2010). Paleoenvironmental reconstruction ofBuluk, Early Miocene, Kenya. Journal of Vertebrate Paleontology,30(Supplemental), 135A.

Morley, C. K., Ngenoh, D. K., & Ego, J. K. (1999a). Introduction to theEast African Rift System. In C. K. Morley (Ed.), Geoscience of the RiftSystems—Evolution of East Africa, AAPG Studies in Geology, 44, 1–18.

Morley, C. K., Wescott, W. A., Stone, D. M., Harper, R. M., Wigger, S.T., Day, R. A., & Karanja, F. M. (1999b). Geology and geophysics of thewestern Turkana Basins. In C. K. Morley (Ed.), Geoscience of the RiftSystems—Evolution of East Africa. AAPG Studies in Geology, 44, 19–54.

Nakaya, H. (1994). Faunal change of Late Miocene Africa and Eurasia:Mammalian fauna from the Namurungule Formation, Samburu Hills,Northern Kenya. African Study Monographs, 20(Supplementary Issue), 1–112.

Nesbit Evans, E. M., Van Couvering, J. A. H., & Andrews, P. (1981).Paleoecology of Miocene sites in Western Kenya. Journal of HumanEvolution, 10, 99–116.

Oksanen, J., Guillaume Blanchet, F., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R. B., Simpson, G. L., Solymos, P., Henry, M., Stevens, H., &Wagner, H. (2013). Vegan: Community Ecology Package. R packageversion 2.0–10. http://CRAN.R-project.org/package=vegan

Ossi, K. M., & Kamilar, J. M. (2006). Environmental and phylogeneticcorrelates of Eulemur behavior and ecology (Primates: Lemuridae).Behavioral Ecology and Sociobiology, 61, 53–64.

Patel, B. A., & Grossman, A. (2006). Dental metric comparisons ofMorotopithecus and Afropithecus: Implications for the validity of thegenus Morotopithecus. Journal of Human Evolution, 51, 506–512.

Peppe, D. J., Deino, A. L., Mcnulty, K. P., Lehmann, T., Harcourt-Smith,W. E. H., Dunsworth, H. M., & Fox, D. L. (2011). New age constraints onthe Early Miocene faunas from Rusinga and Mfangano Islands (LakeVictoria, Kenya). American Journal of Physical Anthropology,144(Supplement 52), 237.

Peppe, D. J., McNulty, K. P., Cote, S. M., Harcourt-Smith, W. E. H.,Dunsworth, H. M., & Vancouvering, J. A. (2009). Stratigraphicinterpretation of the Kulu Formation (Early Miocene, Rusinga Island,Kenya) and its implications for primate evolution. Journal of HumanEvolution, 56, 447–461.

Pickford, M. (1981). Preliminary Miocene mammalian biostratigraphy forwestern Kenya. Journal of Human Evolution, 10, 73–97.

Pickford, M. (1983). Sequence and environments of the Lower and MiddleMiocene Hominoids of Western Kenya. In R. L. Ciochon & R. S.Corruccini (Eds.), New interpretations of ape and human ancestry (pp.421–439). New York: Plenum Press.

Pickford, M. (1987). Fort Ternan (Kenya) palaeoecology. Journal ofHuman Evolution, 16, 305–309.

Pickford, M. (1991). Revision of the Neogene Anthracotheriidae of Africa.The Geology of Libya (pp. 1491–1525). Amsterdam: Elsevier.

Pickford, M. (1995). Fossil land snails of East Africa and theirpalaeoecological significance. Journal of African Earth Sciences, 20, 167–226.

Pickford, M. (2002). New reconstruction of the Moroto hominoid snoutand a reassessment of its affinities to Afropithecus turkanensis. HumanEvolution, 17, 1–19.

Pickford, M., & Andrews, P. (1981). The Tinderet Miocene sequence inKenya. Journal of Human Evolution, 10, 11–33.

Pickford, M., & Kunimatsu, Y. (2005). Catarrhines from the MiddleMiocene (ca. 14.5 Ma) of Kipsaraman, Tugen Hills, Kenya.Anthropological Science, 113, 189–224.

Pickford, M., & Mein, P. (2006). Early Middle Miocene mammals fromMoroto II, Uganda. Beiträge der Paläontologie, 30, 361–386.

Pickford, M., Musalizi, S., Senut, B., Gommery, D., & Musiime, E.(2010). Small apes from the Early Miocene of Napak, Uganda. Geo-PalUganda, 3, 1–110.

Pickford, M., Sawada, Y., Tayama, R., Matsuda, Y.-K., Itaya, T., Hyodo,H., & Senut, B. (2006). Refinement of the age of the Middle Miocene FortTernan Beds, Western Kenya, and its implications for Old Worldbiochronology. Comptes Rendus Geoscience, 338, 545–555.

Pickford, M., Senut, B., Gommery, D., & Musiime, E. (2003). Newcatarrhine fossils from Moroto II, early Middle Miocene (ca 17.5 Ma)Uganda. Comptes Rendues Paleoevolution, 2, 649–662.

Pilbeam, D. & Walker, A. (1968). Fossil Monkeys from Miocene ofNapak North-East Uganda. Nature, 220:657.

Preston, F. W. (1960). Time and space and the variation of species.Ecology 41:611-627.

R Core Team. (2014). R: A language and environment for statisticalcomputing. Vienna, Austria: R Foundation for Statistical Computing.http://www.R-project.org/ .

Reed, K. E. (1997). Early hominid evolution and ecological changethrough the African Plio-Pleistocene. Journal of Human Evolution, 32,

289–322.

Retallack, G. J. (1991). Miocene paleosols and ape habitats of Pakistanand Kenya. New York: Oxford University Press.

Retallack, G. J. (1992a). Middle Miocene fossil plants from Fort Ternan(Kenya) and evolution of African grasslands. Paleobiology, 18, 383–400.

Retallack, G. J. (1992b). Comment on the paleoenvironment ofKenyapithecus at Fort Ternan. Journal of Human Evolution, 23, 363–369.

Retallack, G. J., Wynn, J. G., Benefit, B. R., & McCrossin, M. L. (2002).Paleosols and paleoenvironments of the Middle Miocene, MabokoFormation, Kenya. Journal of Human Evolution, 42, 659–703.

Rohlf, F. J. (1970). Adaptive hierarchical clustering schemes. SystematicZoology, 19, 58–82.

Sanders, W. J., Gheerbrant, E., Harris, J. M., Saegusa, H., & Delmer, C.(2010). Proboscidea. In L. Werdelin & W. J. Sanders (Eds.), Cenozoicmammals of Africa (pp. 161–252). Berkeley: University of CaliforniaPress.

Savage, R. J. G., & Williamson, P. G. (1978). The early history of theTurkana Depression. In W. W. Bishop (Ed.), Geological background tofossil man (pp. 375–394). Edinburgh: Scottish Academic Press.

Senut, B., Pickford, M., Gommery, D., & Kunimatsu, Y. (2000). Unnouveau genre d’Hominoide du Miocene Inferieur d’Afrique Orientale:Ugandapithecus major (Le Gros Clarke & Leakey, 1950). ComptesRendues des Academie des Seances, 331, 227–233.

Shipman, P. (1986). Paleoecology of Fort Ternan reconsidered. Journal ofHuman Evolution, 15, 193–204.

Shipman, P., Walker, A., Van Couvering, J. A., Hooker, P. J., & Miller, J.A. (1981). The Fort Ternan hominoid site, Kenya: Geology, age,

taphonomy and paleoecology. Journal of Human Evolution, 10, 49–72.

Simons, E. L., & Delson, E. (1978). Cercopithecidae and Parapithecidae.In V. M. Maglio & H. B. S. Cooke (Eds.), Evolution of African mammals(pp. 100–119). Cambridge, MA: Harvard University Press.

Simpson, G. G. (1960). Notes on the measurement of faunal resemblance.American Journal of Science, 258A, 300–311.

Smouse, P., Long, J., & Sokal, R. R. (1986). Multiple regression andcorrelation extensions of the Mantel test of matrix correspondence.Systematic Zoology, 35, 627–632.

Soininen, J., McDonald, R., & Hillebrand, H. (2007). The distance decayof similarity in ecological communities. Ecography, 30, 3–12.

Sokal, R. R., & Sneath, P. H. A. (1963). Principles of numericaltaxonomy. San Francisco: W. H. Freeman.

AQ13

SPSS Inc. (Released 2009). PASW Statistics for Windows, Version 18.0.Chicago: SPSS Inc.

Stevens, N. J., Seiffert, E. R., O’Connor, P. M., Roberts, E. M., Schmitz,M. D., Krause, C., Gorscak, E., Ngasala, S., Hieronymus, T. L., & Temu,J. (2013). Palaeontological evidence for an Oligocene divergence betweenOld World monkeys and apes. Nature, 497, 611–614.

Tsubamoto, T., Takai, M., & Egi, N. (2004). Quantitative analyses ofbiogeography and faunal evolution of Middle to Late Eocene mammals inEast Asia. Journal of Vertebrate Paleontology, 24, 657–667.

Ungar, P., Scott, J., Curran, S., Dunsworth, H., Harcourt-Smith, W.,Lehmann, T., Manthi, F., & McNulty, K. (2012). Early Neogeneenvironments in East Africa: Evidence from dental microwear of tragulids.Palaeogeography, Palaeoclimatology, Palaeoecology, 342, 84–96.

Van Couvering, J. A. H., & Van Couvering, J. A. (1976). Early Miocenemammal fossils from East Africa: Aspects of geology, faunistics andpaleo-ecology. In G. L. Isaac & E. R. McCown (Eds.), Human origins:Louis Leakey and the East African evidence (pp. 155–208). Menlo Park,CA: W. A. Benjamin.

Van Damme, D., & Gautier, A. (1972). Some fossil molluscs fromMuruarot Hill (Turkana District, Kenya). Journal of Conch, 27, 423–426.

Wessels, W., Fejfar, O., Peláez-Campomanes, P., van der Meulen, A., &de Bruijn, H. (2003). Miocene small mammals from Jebel Zelten, Libya.Coloquios de Paleontología, Ext., 1, 699–715.